Therapeutic antibodies constitute a major class of current drugs under development because they can bind with high affinity and specificity to particular targets in the body or on an infectious agent. The structure of a therapeutic antibody bound to its antigen can be of value to reveal the structural origin of the mechanism of the drug's action and possibly indications of the mechanism of the antigen itself in a disease process. Also, the structure can be exploited for further protein engineering such as increasing binding affinity. We propose to develop tools to predict the structure of antibody-antigen complexes starting from the sequence of the antibody and an unbound structure of the antigen. We will improve and tailor docking scoring functions, develop homology models for docking, and develop flexible loop algorithms for docking to account for uncertainties in a homology antibody structure. Antibodies make an ideal test system for homology docking and flexible loop docking because the IgG fold is well-defined;when these techniques are developed, they will be of broad impact for the docking field in general. Finally, we will apply these techniques to two model systems for which the antibody-antigen complex structure is unknown. Monoclonal antibody 806 binds to the epidermal growth factor receptor and has been shown to be an effective therapy against several types of tumors. The 14B7 family of antibodies binds the anthrax toxin and prevents its cytotoxic effects. Both complex structures are currently unknown. We will predict structures and collaborate to validate the predictions experimentally. This research is relevant to public health because it will provide general computational tools to discover the structural mechanism of new protein drugs. In addition, this research will determine the structure for a novel cancer drug bound to its target and a novel antidote bound to the anthrax toxin, revealing the structural basis of the therapies and providing a foundation to tailor the drugs to increase their effectiveness.